All-in-one medical image-to-image translation

The growing availability of public multi-domain medical image datasets enables training omnipotent image-to-image (I2I) translation models. However, integrating diverse protocols poses challenges in domain encoding and scalability. Therefore, we propose the "every domain all at once" I2I (EVA-I2I) translation model using DICOM-tag-informed contrastive language-image pre-training (DCLIP). DCLIP maps natural language scan descriptions into a common latent space, offering richer representations than traditional one-hot encoding. We develop the model using seven public datasets with 27,950 scans (3D volumes) for the brain, breast, abdomen, and pelvis. Experimental results show that our EVA-I2I can synthesize every seen domain at once with a single training session and achieve excellent image quality on different I2I translation tasks. Results for downstream applications (e.g., registration, classification, and segmentation) demonstrate that EVA-I2I can be directly applied to domain adaptation on external datasets without fine-tuning and that it also enables the potential for zero-shot domain adaptation for never-before-seen domains.

Authors

Institutions

Publication Details

Journal
Cell Reports Methods
Published
2025-08-01
DOI
https://doi.org/10.1016/j.crmeth.2025.101138
Citations
2
Primary Topic
Multimodal Machine Learning Applications
Type
article
Field-Weighted Citation Impact
2.55

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

All-in-one medical image-to-image translation

Chunyao Lu, Luyi Han, Shubo Zhou, Yu-E Sun et al.
2 citations
Cell Reports Methods
Multimodal Machine Learning Applications
2.55
article

All-in-one medical image-to-image translation

Chunyao Lu, Luyi Han, Shubo Zhou, Yu-E Sun, Yunzhi Huang, Ritse M. Mann, Haoran Dou, Xin Wang, Tianyu Zhang, Yuan Gao, Xinglong Liang, Tao Tan, Jonas Teuwen
article en
2 citations

Abstract

The growing availability of public multi-domain medical image datasets enables training omnipotent image-to-image (I2I) translation models. However, integrating diverse protocols poses challenges in domain encoding and scalability. Therefore, we propose the "every domain all at once" I2I (EVA-I2I) translation model using DICOM-tag-informed contrastive language-image pre-training (DCLIP). DCLIP maps natural language scan descriptions into a common latent space, offering richer representations than traditional one-hot encoding. We develop the model using seven public datasets with 27,950 scans (3D volumes) for the brain, breast, abdomen, and pelvis. Experimental results show that our EVA-I2I can synthesize every seen domain at once with a single training session and achieve excellent image quality on different I2I translation tasks. Results for downstream applications (e.g., registration, classification, and segmentation) demonstrate that EVA-I2I can be directly applied to domain adaptation on external datasets without fine-tuning and that it also enables the potential for zero-shot domain adaptation for never-before-seen domains.

Cell Reports MethodsVol. 5(8)
University of Leeds (GB), Radboud University Nijmegen (NL), Nanjing University of Information Science and Technology (CN), Maastricht University Medical Centre (NL), Radboud University Medical Center (NL), The Netherlands Cancer Institute (NL), Institute of Computing Technology (CN), Suzhou Institute of Biomedical Engineering and Technology (CN), Macao Polytechnic University (MO)
China Scholarship Council
Openalex Percentile: Top 9%
Multimodal Machine Learning Applications
2.55
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.